How to spot a honeypot token before you buy

Updated 2 September 2026 · Belphor Research · the data behind this guide

A honeypot is a market with a door that only opens inward: buying works perfectly, selling fails or is taxed into confiscation. The chart keeps rising because nobody who bought can sell, which makes honeypots look like winners right up until you own one. The defense is one habit: test the exit before you enter.

Why the chart cannot warn you

Every other scam eventually shows up in the price. A honeypot is the exception: because selling is disabled, the only trades that exist are buys, so the chart shows a clean uptrend with no red candles worth noticing. Volume looks healthy, holders grow, and social proof accumulates from buyers who do not yet know they are locked in. The signal you would normally trust, price action, is manufactured by the mechanism itself. That is what makes honeypots the most seductive trap in the market.

The mechanisms

Implementation varies. Some tokens carry transfer logic that blocks sales from any wallet not on an allowlist. Some maintain a blacklist the operator adds buyers to after they enter. Some set a sell tax of ninety-plus percent, technically letting you sell while keeping the proceeds. And some build the trap into the pool rather than the token, with liquidity structured so that sell routes fail. On Solana, token-level programmability is narrower than on EVM chains, which pushes traps toward the pool and tax styles, but the effect is identical: your money enters and cannot leave.

The sell-route test

The single most effective check costs nothing: before buying, ask a swap aggregator to quote the reverse trade, selling the token for SOL. If no route comes back while the buy route works fine, you have found a one-way market, and no further analysis is needed. This is the test Belphor wired directly into its live trading engine: every real entry first quotes the exact tokens it is about to receive back to SOL, and refuses the trade if no route exists.

The round-trip cost check

The subtler honeypots pass the route test but tax the exit. These are exposed by comparing the two quotes: what you pay going in versus what the just-quoted position would fetch coming straight back out. Normal thin-market friction costs a few percent; a round-trip loss of fifteen, thirty, or eighty percent means the difference is being confiscated by design. Set a threshold and make it mechanical. Belphor refuses live entries whose round-trip loss exceeds its cap, no matter how good the signal looks, because a profit you cannot withdraw is not a profit.

The supporting cast

Honeypots correlate with the rest of the rug toolkit, so the standard checks still apply: unrevoked authorities, unlocked LP, extreme holder concentration, and a deployer wallet with history. The full list lives in the anatomy of a rug pull. And one behavioral rule beats everything else: when a coin you have never heard of shows a perfect chart and an urgent community, the correct first act is to test selling, not to buy. Traps are built for people who skip that step.